EARLY TESTING AND RECURRENT SELECTION 417 



safest and most efficient method to use with our j)resent lack of knowledge 

 concerning the relative imjiortance of partial dominance, dominance, and 

 over-dominance in determining combining ability. 



In the discussion presented so far no emphasis has been placed upon choice 

 of testers. It is obvious that either early testing or recurrent selection can be 

 carried out giving special emphasis to either general or specific combining 

 ability depending upon the tester parent chosen. In the experiments involv- 

 ing oil percentage of the grain this problem does not arise. In the e.xperiments 

 involving test crosses for yield evaluation, double crosses or open-pollinated 

 varieties have been used as tester parents thus giving special emphasis to 

 general combining ability. 



SUMMARY 



In the data which have been presented bearing on early testing, the 

 method has demonstrated all of the characteristics which have been claimed 

 for it. This is not to be interpreted as meaning early testing is the ideal corn 

 breeding method and equally applicable under all circumstances. It is useful 

 under some conditions. The ideal method of corn breeding probably is still 

 to be devised. 



Recurrent selection has been found to be quite effective in modifying the 

 chemical composition of the corn grain. Tests of this method in modifying 

 combining ability have been less extensive. Here again this method may not 

 be equally valuable under all conditions and circumstances, but on the basis 

 of results to date it is certainly deserving of more extensive use. 



